Children and Youth Services Review 31 (2009) 410421

Contents lists available at ScienceDirect

Children and Youth Services Review

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h i l d yo u t h

Fostering security? A meta-analysis of attachment in adopted children

Linda van den Dries, Femmie Juffer , Marinus H. van IJzendoorn, Marian J. Bakermans-Kranenburg
Centre for Child and Family Studies, Leiden University, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Adopted children are hypothesized to be at risk of insecure attachment relationships because of their
Received 30 May 2008 background of institutional care, maltreatment and neglect. We conducted two series of meta-analyses, one
Received in revised form 8 September 2008 using only observational assessments of attachment and one using both observational and self-report
Accepted 14 September 2008
assessments. Observational assessments showed that children who were adopted before 12 months of age
Available online 24 September 2008
were as securely attached as their non-adopted peers, whereas children adopted after their rst birthday
Keywords:
showed less attachment security than non-adopted children (d = 0.80, CI = 0.491.12). Regarding the overall
Meta-analysis effect for attachment security, adoptees were comparable to foster children. Adopted children showed more
Adoption disorganized attachments compared to their non-adopted peers (trimmed d = 0.36, CI = 0.040.68), but again
Attachment were comparable to foster children (trimmed d = 0.35, CI = 0.020.67). Compared to institutionalized children,
Parentchild relationships adoptees were less often disorganized attached. When self-report measures of attachment were included no
Institutional care difference was found between adoptees and their non-adopted counterparts (trimmed d = 0.12, CI = 0.02
Foster care 0.26, 39 studies, N = 2912 adopted children). Compared to institutionalized children, (early) adoption proves
to be an effective intervention in the domain of attachment.
2008 Elsevier Ltd. All rights reserved.

1. Introduction
Are adopted children less often securely attached to their adoptive
parents than children reared by their biological parents? In a series of
meta-analyses we examined adopted children's attachment relation-
ships with their adoptive parents. Previous meta-analyses on the
development of adoptees documented catch-up after adoption in various
domains, for example in cognitive development (Van IJzendoorn, Juffer,
& Klein Poelhuis, 2005) and physical growth (Van IJzendoorn, Baker-
mans-Kranenburg, & Juffer, 2007), but less so for learning problems (Van
IJzendoorn et al., 2005) and clinical referrals (Juffer & Van IJzendoorn,
2005). The question to be addressed here is whether adopted children
also show catch-up or delays in the domain of attachment relationships
after the transition to their new families. As adopted children's
characteristics and experiences show some similarities to those of foster
children, we compared the outcomes for adopted children not only with
This study was funded by the Netherlands Organization for Scientic Research
(NWO 400-03-208), and additional support was received from VSBfonds, Fonds
Psychische Gezondheid, and Stichting Kinderpostzegels Nederland to Femmie Juffer
and Marinus van IJzendoorn in cooperation with the Adoptie Driehoek Onderzoeks
Centrum [Adoption Triad Research Center] (www.adoptionresearch.nl). Femmie Juffer
is supported by Wereldkinderen. Marinus van IJzendoorn and Marian Bakermans
Kranenburg are supported by the Netherlands Organization for Scientic Research
(SPINOZA Prize and VIDI grant, respectively).
Corresponding author. Leiden University, Centre for Child and Family Studies, P.O.
Box 9555, 2300 RB Leiden, The Netherlands.
E-mail address: juffer@fsw.leidenuniv.nl (F. Juffer).
children from biological families but also with outcomes for foster
children.
Bowlby (1982, p. 371) stated about the nature of attachment
relationships: To say of a child that he () has an attachment to
someone means that he is strongly disposed to seek proximity to and
contact with a specic gure and to do so in certain situations, notably
when he is frightened, tired or ill. Although the tendency to form
attachment relationships is innate and universal, individual differ-
ences can be observed in the quality of attachment. Based on the
attachment strategies children use when they face stressful situations,
their attachment relationships can be classied as secure, insecure
(avoidant or ambivalent) or insecure-disorganized. Secure children
seek contact with their attachment gure when they are upset and are
easily comforted. Insecure children, on the other hand, show signs of
avoidance or resistance (Ainsworth, Blehar, Waters, & Wall, 1978).
Disorganized attachment is considered the most insecure type of
attachment, with disorganized children showing a breakdown of a
consistent attachment strategy when dealing with a stressful situation
(Main & Hesse, 1990).
Developing a secure attachment relationship or close bond with a
parent or primary caregiver has long-term benets for children,
because of the impact on children's later adaptation, for instance
through the development of emotion regulation (Cassidy & Shaver,
1999). Secure infants develop basic trust in their parents and they feel
condent about their own ability to inuence the world around them.
This basic trust allows children to function autonomously and with
condence in their (social) problem solving abilities (Sroufe, Egeland,
Carlson, & Collins, 2005; Weineld, Sroufe, Egeland, & Carlson, 1999).

0190-7409/$ see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.childyouth.2008.09.008
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
Insecurely attached children are at risk for a more problematic
development. For example, insecure attachment, in particular inse-
cure disorganized attachment, has been associated with the develop-
ment of externalizing behavior problems (e.g., Lyons-Ruth,
Easterbrooks, & Cibelli, 1997) and subsequent child psychopathology
(for a meta-analysis see Van IJzendoorn, Schuengel, & Bakermans-
Kranenburg, 1999).
2. Adoption and the comparison with foster children
Every year, large numbers of children are placed in adoptive
homes. In 2004 almost 45,000 children were adopted internationally
worldwide (Selman, 2006). The number of domestic adoptions is
difcult to estimate, as not all of these adoptions are ofcially
registered (Placek, 2007). In the USA, 1.6 million children live with
adoptive parents (Jones, 2008) and between 2 to 4 percent of the
families include an adopted child. Annually about 20,000 children are
adopted internationally in the USA (Brooks, Simmel, Wind, & Barth,
2005; Nickman et al., 2005) and 50,000 children are domestically
adopted from the public system, whereas the practice of relinquishing
an infant for domestic adoption is estimated at about 1% of babies born
to never-married women (Jones, 2008). In Western-European coun-
tries, with the exception of the United Kingdom and Germany, there
are relatively few domestic adoptions compared to the number of
international adoptions (Selman, 2006), while adoptions from foster
care are rare, again with the United Kingdom as an exception
(Warman & Roberts, 2003).
In the USA there were 510,000 children in foster care in 2006
(Adoption and Foster Care Analysis and Reporting System, 2008).
Because foster children experience comparable disruptions of early
parentchild relationships as adopted children, we also compared the
outcomes of adopted children with the outcomes of foster children.
The background of adopted children shows some similarities with the
background of children placed in foster homes. Both groups have
experienced separation from their birth parents, and are placed in
new families where they are cared for by foster or adoptive parents
who usually show little (physical) resemblance to them. In addition,
many adopted and foster children have experienced unfavorable
conditions before placement. The majority of internationally adopted
children have experienced neglect and maltreatment in institutional
care before adoption (Miller, 2005; Tirella et al., 2008). In the same
vein, many foster children have been the victim of neglect and/or
repeated abuse before they enter the foster care system, as was
conrmed in several large scale studies (Chernoff, Combs-Orme,
Risley-Curtiss, & Heisler, 1994; Department for Children, Schools and
Families, 2008; Lewit, 1993; Takayama, Wolfe, & Coulter, 1998),
although not all children have experienced the same level of adversity
before foster placement (e.g., Quinton, Rushton, Dance, & Mayes,
1998). Domestic adoptees may even resemble foster children more, as
some domestic adoptees are former foster children (Testa, 2004).
Nevertheless, there are also several marked differences between
adopted and foster children. One of these differences is the
permanency of the placement. Whereas the placements of adopted
children are permanent, placements of foster children are often
temporary. Adopted and foster children may also be placed out of their
homes for different reasons. International adoptees are usually placed
for cultural reasons, extreme poverty or family policy (e.g., the one-
child policy in China; Johnson, 2004), while foster children for
example are placed because of caretaker incapacity or absence, child
protective reasons, or parental incarceration (Chernoff et al., 1994;
Hayward & DePanlis, 2007; Lewit, 1993). In addition, whereas
adoption is a clear-cut situation in which children have been removed
from their birth parents and, except in open adoption arrangement,
have no contact with the birth parents, there is a high diversity of
foster care arrangements, many implying contact and visits with the
biological parent (Leathers, 2003; Sanchirico & Jablonka, 2000).
411
3. Attachment in adopted children
In some studies adopted children have been reported to show
reactive attachment disorders (Zeanah, 2000; Zeanah et al., 2004) and
indiscriminate friendly behavior (Chisholm, Carter, Ames, & Morison,
1995; Tizard & Rees, 1975). More insecure and disorganized attach-
ments (Marcovitch et al., 1997) and non-optimal parentchild
relationships (Fletcher, 1995) have been found in adoptees as well.
There are several reasons to expect less attachment security in
adopted children, as these children all experience separation from,
and loss of their birth parents and other caregivers. According to
attachment theory, these experiences of loss and separation may
negatively inuence the development of subsequent attachment
relationships (Bowlby, 1982). In addition, many internationally
adopted children have been in institutional care and have experienced
deprivation, neglect and maltreatment. Children living in institutions
often lack opportunities to develop selective attachment relation-
ships, due to the limited amount and poor quality of contact with their
caregivers (Gunnar, Bruce, & Grotevant, 2000; Howe, 2005; Palacios &
Snchez-Sandoval, 2005; Vorria et al., 2003). The debate about the
effects of early deprivation on child development dates back more
than half a century ago, when the World Health Organization initiated
a study conducted by child psychiatrist John Bowlby. In a now famous
report, Bowlby (1952) concluded that children suffered from the
effects of institutional care, even when their physical needs (food,
clothes, etc.) were met adequately. The children were deprived of
parental care and missed out opportunities to develop stable and
continuous attachment relationships. According to Bowlby, early
deprivation leads to compromised child development and sets the
stage for various mental health problems in children. As viable
alternatives for institutional care Bowlby (1952, p.109) recommended
adoption and foster care, because they provide children with
substitute parents.
As adoption implies separations, loss, and the development of
attachment relationships to new parents, attachment theory (Bowlby,
1982) provides an appropriate theoretical framework for our study. In
addition, adoption also implies risks (e.g., deprivation before place-
ment) as well as protection (e.g., receiving nurturing care and
stimulation from alternative parent gures) and therefore the
perspective of risk and protective factors is also important. According
to this theoretical perspective, an accumulation of risk factors leads to
less optimal child development, whereas protective factors may buffer
the negative effects of the risks, resulting in resilience in children
(Rutter, 1990; Werner, 2000).
Recent neurobiological studies of institutional care suggest that
these high stress environments inuence brain development and
attachment behaviors and may cause persistent cognitive and socio-
emotional delays (Miller, 2005; Rutter, 2005, 2006; Rutter et al.,
2004). Studies showing altered patterns of cortisol (Gunnar, Morison,
Chisholm, & Schuder, 2001), neurocognitive impairment (Chugani et
al., 2001), and changes in the production of neuropeptides (Fries,
Ziegler, Kurian, Jacoris, & Pollak, 2005) after institutional care point to
a critical role for early experience in the development of the brain
systems underlying basic aspects of human social behavior and stress
regulation. According to Gunnar and Kertes (2005) adverse experi-
ences may affect the structure and function of the brain in three
general ways. First, severe malnutrition and maltreatment can injure
neural tissue; second, a lack of stimulation can affect the basic wiring
plan of the brain and brain chemistry; and third, morphological and
neurochemical adaptations to a non-optimal (institutional) environ-
ment may produce maladaptive responses to the post-institutional
environment that limit (later) healthy behavioral and emotional
development (Gunnar & Kertes, 2005, p. 49).
Two studies conrmed the disadvantageous effects of institutional
care on attachment, with institutionalized children showing high
rates of insecure attachment and especially high rates of disorganized
412 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
attachment (Vorria et al., 2003; Zeanah, Smyke, Koga, Carlson, & the
BEIP Core Group, 2005). Although domestically adopted children not
always experience institutional care, the neglect and maltreatment
they often face in their birth families may also have a detrimental
effect (Kaniuk, Steele, & Hodges, 2004). In particular maltreatment is a
documented precursor of disorganized attachment (Carlson, Cicchetti,
Barnett, & Braunwald, 1989; Crittenden & Ainsworth, 1989; George,
1996; Van IJzendoorn et al., 1999), while experiences of parental
neglect are reported to be a precursor of insecure (ambivalent)
attachments (Egeland & Sroufe, 1981; Finzi, Ram, Har-Even, Shnit, &
Weizman, 2001; Youngblade & Belsky, 1990).
Almost all children develop an affective bond with their caregiver,
even maltreated children (Cicchetti & Barnett, 1991) or children
struggling with autism (Rutgers, Van IJzendoorn, Bakermans-Kranen-
burg, & Swinkels, 2007). Based on the caregiver's reactions to their
signals, children develop expectations (so-called internal working
models) of how they will be treated by their attachment gures
(Bowlby, 1982; Bretherton & Munholland, 1999). Children whose
attachment signals are met by sensitive caregivers develop an internal
working model of a safe and responsive world. In contrast, children
who are responded to in an insensitive way may picture the world as
an unpredictable place and they may not feel worthy of love.
Children's internal working models are suggested to inuence the
development of new relationships, by shaping children's expectations
and behavior (Bowlby, 1982; Sroufe et al., 2005). But Bowlby also
hypothesized that working models can change as a consequence of
changing experiences, in particular in the rst ve years of life, and
that corrective attachment experiences may compensate for early
adversity (Bowlby, 1988). When transitions to responsive care are
experienced, internal working models and attachment relationships
may change correspondingly (Bowlby, 1973, 1988; Sroufe et al., 2005).
Whether this process of accommodation also takes place after the
placement of children with (sensitive) adoptive parents is a central
question in the current set of meta-analyses. Research on attachment
in adopted children shows equivocal outcomes. In some studies
attachment of adopted children does not differ from the security of
non-adopted children (e.g., Joseph, 2002) whereas other studies nd
fewer secure attachments in adopted children (e.g., O'Connor, Marvin,
Rutter, Olrick, & Britner, 2003). With the current meta-analysis we
attempted to answer the question whether adopted children show
less attachment security than non-adopted children. We also tested
whether adopted children show more disorganized attachment than
non-adopted children.
4. Potential factors inuencing attachment relationships in
adoptive families
In this meta-analysis the inuence of various moderators has
been investigated, namely, age at placement, duration of placement,
continent of origin, domestic or international placement, and trans-
racial or same-race placement. From the perspective of risk and
protective factors some moderators may imply risks (e.g., an older age
at placement) or protection (e.g., a longer stay in the adoptive home).
In particular, age at placement may be crucial for the development of a
new relationship with the adoptive parent. When children receive
warm and sensitive care, they generally develop basic trust in their
caregivers during their rst year of life (Bowlby, 1982). Several studies
indeed found that children who were adopted in the rst months of
their lives, usually develop normative attachment relationships (e.g.,
Juffer, Bakermans-Kranenburg, & Van IJzendoorn, 2008), whereas
children who were placed at later ages seem to be at greater risk for
developing unfavorable attachment relationships (Marcovitch et al.,
1997; Vorria et al., 2006).
The length of time the children have spent in their new family may
also be a signicant moderator. Children who have lived with their
new parents for a longer period of time, and thus have spent more
time in the care of a stable and usually nurturing parent, may have had
more time to recover from prior adverse experiences. In a similar vein,
Juffer and Van IJzendoorn (2005) meta-analytically found that
children who had spent more than 12 years in their adoptive family
showed a larger catch-up in terms of behavior problems than children
who had lived in their adoptive homes for a shorter period of time.
In addition, continent of origin may be a relevant risk factor to take
into account when explaining differences in adaptation. Children who
originate from Eastern European countries, for example Romania,
often have experienced severe deprivation (Castle et al., 1999; Miller,
2005; Morison, Ames, & Chisholm, 1995; Smyke et al., 2007), and may
therefore show more problems with attachment than children
adopted from other continents. Similar results were found in previous
meta-analytic work: children who experienced more severe depriva-
tion showed more behavior problems and lower cognitive compe-
tence than children from less deprived backgrounds (Juffer & Van
IJzendoorn, 2005; Van IJzendoorn et al., 2005).
Finally, type of placement may be an important risk factor. In some
domains, for example, problem behavior and mental health referrals,
differences have been reported between international and domestic
adoptees (Juffer & Van IJzendoorn, 2005), with international adoptees
showing fewer behavior problems and mental health referrals.
However, for other developmental domains, such as self-esteem, no
differences have been found between international and domestic
adoptees, or between same-race or transracial adoptees (Juffer & Van
IJzendoorn, 2007; Van IJzendoorn et al., 2005). With regard to
attachment we did not expect to nd differences between the above
mentioned groups, because early experiences of separation and loss,
and/or neglect and maltreatment are assumed to be present in the
majority of the adopted children.
Based on the literature presented above, we hypothesized that
adopted children would show fewer secure and more disorganized
attachments compared to children living with their biological parents.
We expected that these differences would be larger in children who are
adopted after their rst birthday than in children adopted in their rst
year of life. We also hypothesized that children who have lived with their
new parents for a longer period of time will show a more favorable
relationship with the parent. Finally, we expected children who are born
in Eastern European countries to show less secure and more dis-
organized attachment than children adopted from other continents.
5. Method
5.1. Literature search
To identify relevant studies three different search methods were
used. First, we searched for relevant literature in the following
electronic sources: PsycInfo (Psychological Literature), ERIC (Educa-
tion Resource Information Center), Web of Science and PUBMED (U.S.
National Library of Medicine). Throughout this search we used the
keywords adopt or foster (an asterisk indicates that the search
contained but was not limited to that word or word fragment) in
combination with the terms attachment, parentchild relationship,
bonding, and related terms such as security, motherchild relationship,
Strange Situation and AQS. Second, we searched the reference lists of
all collected studies to identify more relevant studies. Third, experts
were asked to provide pertinent studies.
Studies were included if they reported on the attachment re-
lationship, parentchild relationship or bond between the adopted or
foster children and their adoptive/foster parents (hereafter: adopted
children and adoptive parents). Both studies using observational
assessments and self-report or parent-report measures were included.
In order to examine the attachment relationships of the adoptees
across their lifespan, no restriction was placed on the age at assess-
ment. In the case of a study with more than one assessment (e.g., Juffer
& Rosenboom, 1997), only the data from the rst valid assessment was
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
included to guarantee that every adoptee was included only once in a
meta-analysis. Studies measuring attachment relationships were
excluded if: (1) the adopted sample was a clinical sample (e.g.,
Rosenthal et al., 1975); (2) the relationship between the adoptee and
someone else than the caregiver was measured (e.g., Jensen, 2004);
(3) the information in the study was not sufcient to compute an
effect size (e.g., Judge, 2004); (4) attachment style was measured (e.g.,
Borders, Penny, & Portnoy, 2000) or (5) no comparison group was
available (e.g., Brown, 2000; Hodges, Steele, Hillman, Henderson, &
Kaniuk, 2005). The presence of a comparison group in the study was
not required when studies reported on the distribution of attachment
classication based on the (adapted) Strange Situation Procedure (see
below), the Attachment Q-sort (Verssimo & Salvaterra, 2006), or
applied instruments for which a comparison group of another study
could be used (e.g., Millham, 2003). We excluded studies measuring
attachment style (e.g. Borders et al., 2000), since our study focuses on
the attachment relationships of the adoptees with their adoptive
parents, instead of on their romantic or intimate attachment relation-
ships in general.
Attachment relationships can be measured using various instru-
ments. The Strange Situation Procedure (SSP; Ainsworth et al., 1978) is
an observational laboratory procedure used to assess infant attach-
ment behavior. The SSP consists of eight 3-minute episodes during
which two separations from, and reunions with, the parent occur. The
attachment of the children in the SSP is classied based on the (in-)
secure patterns of attachment behavior and the presence of
disorganized attachment behavior. Some studies use an adapted SSP,
with coding systems such as the CassidyMarvin system, the Main
Cassidy system, and Crittenden's Preschool Assessment of Attachment
(PAA). As the traditional Ainsworth classications can only be used in
children up to 18 months of age, Cassidy and Marvin (1992) developed
a classication system for preschool-age children. Crittenden (1992)
did the same with her classication system and Main and Cassidy
(1988) developed a classications system for kindergarten-age
children. These approaches use a procedure comparable to the
Strange Situation Procedure, namely one or two separations and
reunions (Solomon & George, 1999). The Attachment Q-Sort (AQS;
Waters & Deane, 1985) is another observational measure used to
assess attachment security. The AQS consists of 90 items, each
describing specic behavioral characteristics of the child with an
emphasis on secure-base behavior. After several hours of observation
an observer sorts all cards into nine piles of 10 cards each, depending
on how well the description ts the child. By comparing the child's
prole with the behavioral prole of a prototypically secure child, a
score for attachment security can be derived. The SSP and AQS are
widely used and meta-analytically validated observational instru-
ments for assessing attachment in infants and toddlers/preschoolers
(Solomon & George, 1999; Van IJzendoorn, Vereijken, Bakermans-
Kranenburg, & Riksen-Walraven, 2004). The CassidyMarvin system,
the MainCassidy system and Crittenden's PAA all have been found to
be valid assessment instruments (Moss, Bureau, Cyr, Mongeau, & St-
Laurent, 2004; Main & Cassidy, 1988; Teti & Gelfand, 1997), although
some questions about the concordance among the systems have been
raised as well (Crittenden, Claussen, & Kozlowska, 2007; Solomon &
George, 1999). As part of our meta-analyses, we will therefore test
whether including these measures will affect the results.
As the observational attachment measures have the best creden-
tials in terms of validation, we were particularly interested in the
outcomes of studies using the (adapted) SSP or the AQS: in our
analyses they constitute the core set of studies. To compare the effect
sizes in this core set of studies on adopted children with studies on
foster children, a separate set of meta-analyses for foster children was
conducted, including again only studies that used the (adapted) SSP or
the AQS. However, as attachment relationships of adopted children
have also been assessed using questionnaires or projective measures
we repeated the meta-analysis on adopted children using all types of
413
measures in order to see if results converged (see Table 1 for all
included studies and measures). One example of such a questionnaire
is the Inventory of Parent and Peer Attachment (IPPA; Armsden &
Greenberg, 1987). The IPPA consists of 28 items concerning the
behavioral and affective/cognitive dimension of an adolescent's
attachment relationships with their parents (Armsden & Greenberg,
1987). The IPPA was used in several studies with adopted children, e.g.,
the studies of Fischman (1996) and McGinn (2001).
For studies using the (adapted) SSP the distribution of (in-)security
in the adopted group was compared to the normative distribution of
ABC-classications reported in the meta-analysis of Van IJzendoorn,
Goldberg, Kroonenberg and Frenkl (1992; k = 21, N = 1584). Studies
reporting on ABCD-classications were compared with the normative
distribution reported in the meta-analysis of Van IJzendoorn et al.
(1999) for the distributions of (in-)security and (dis-)organization
(k = 15, N = 2104). The mean security score of the AQS (M = .32, k = 28,
N = 2516) from the meta-analysis of Van IJzendoorn et al. (2004) was
used as a comparison for the studies using the AQS. The normative
attachment scores presented in these meta-analyses were based on
large samples and can therefore be seen as more reliable than the
attachment scores of the often small-scale control groups presented in
the studies themselves.
In our series of meta-analyses we drew on 39 adoption and 11
foster studies (reported in 39 publications), with 21 studies using the
(adapted) SSP, seven studies using the AQS, six studies using the
Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979),
four studies using the IPPA (Armsden & Greenberg, 1987), six studies
using a shortened version of the AQS (Chisholm et al., 1995) and six
studies using other measures, such as the Adult Attachment Interview
(AAI; George, Kaplan, & Main, 1985).
Of the 11 foster studies, ve studies examined prenatally drugs-
exposed foster children. To examine whether this affected the
outcomes of the studies, we used prenatal drug exposure as a
moderator in preliminary analyses. Because no difference was found
between the two sets of studies, Q(1) = 0.002, p N .05, the studies with
the prenatally exposed children were included in the meta-analyses
on foster children.
5.2. Data extraction
We conducted several moderator analyses. Given that moderator
analysis is only relevant when the different subsets comprise of more
than three studies each (Bakermans-Kranenburg, Van IJzendoorn, &
Juffer, 2003), subsets that consisted of fewer than four studies were
not included in the contrast analyses. We coded the following study
characteristics: publication outlet, year of publication (b1995, 1995
1999, N2000 or in press) and continent of study (for example, North
America or Europe). With regard to publication outlet, we tested the
contrast between studies published in refereed journals and non-
refereed publications (book, chapter article and presentation). As 13 of
our studies were dissertations, we created a separate group for this
type of publication.
We coded the sample sizes of the adoption and control groups, age
at placement (before or after 12 months of age), age at assessment (0
4 years, 412 years, N12 years), time spent in the new family (0
1 years, 12 years, 212 years, N12 years), continent of origin (Asia,
North America, Europe, other continents or several continents) and
type of placement (international or domestic, and same-race or
transracial). We also examined the inuence of age at assessment. As
we expected that age at placement would be an important variable for
potential catch-up or delay, no specic expectations for age at
assessment were formulated. For example, children who were
assessed at age 4.5 could have been placed at different ages, e.g.,
6 months versus 24 months of age. We expected that the inuence of
age at placement would overrule the inuence of age at assessment. In
the same vein, we expected that the time spent in the new family
414 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421

Table 1
Attachment relationships of adoptees vs. non-adopted comparisons

Source NaNc Age at arrival, m Age at study, y Country of study Measure (informant) Core set (in)secure Core set disorganized
Bartel, 2006a 1239 12 04 USA Shortened version AQS (P) No No
1069 12 04 USA Shortened version AQS (P) No No
Caspers, Yucuis, Troutman, Arndt, 126584 b12 N 18 USA AAI (O) No No
& Langbehn, 2007
Chisholm, 1998b, c 262104 b12 412 Canada SSP, Crittenden (O) Yes Yes
432104 12 412 Canada SSP, Crittenden (O) Yes Yes
Farina, Leifer, & Chasnoff, 2004a 298 12 04 USA Shortened version AQS (P) No No
Feeney, Passmore, & Peterson, 2007 144131 b12 N 18 Australia PBI Care (S) No No
Fischman, 1996 17369 n.r. N 18 USA IPPA (S) No No
Fletcher, 1997 100100 n.r. N 18 USA PBI Care (S) No No
b
Irhammar & Bengtsson, 2004 40584 12 N 18 Sweden AAI (O) No No
Joseph, 2002 3030 12 N 18 USA IPPA Trust (S) No No
Juffer et al., 2005b, c 782104 b12 04 Netherlands SSP (O) Yes Yes
Juffer & Roosenboom, 1997b 302104 b12 04 Netherlands SSP (O) Yes No
Lis, 2000b 81584 12 04 Poland Separation-Reunion (O) No No
81584 12 04 Poland Separation-Reunion (O) No No
b, c
Marcovitch et al., 1997 442104 12 04 Canada SSP, CassidyMarvin (O) Yes Yes
McGinn, 2001 3030 b12 1218 USA IPPA Attachment (S) No No
Millham, 2003a 168 b12 04 USA Shortened version AQS (P) No No
158 b12 412 USA Shortened version AQS (P) No No
Mller, Gibbs, & Ariely, 2002b 33070 b12 n.r. USA IPPA Attachment (S) No No
O'Connor et al., 2003b, c 432104 b12 412 UK SSP, CassidyMarvin (O) Yes Yes
492104 b12 412 UK SSP, CassidyMarvin (O) Yes Yes
392104 12 412 UK SSP, CassidyMarvin (O) Yes Yes
b, c
Ongari & Tomasi, 2006 62104 b12 04 Italy SSP, CassidyMarvin (O) Yes Yes
Pace, Messina, Zavattini, & 112104 12 412 Italy SSP, MainCassidy (O) Yes No
Santona, 2006b
Paperny, 2004b 34487 b12 n.r. Canada Adult Attachment Projective (S) No No
Passmore et al., 2005 5050 b12 n.r. Australia PBI Care (S) No No
5050 b12 n.r. Australia PBI Care (S) No No
Rosnati & Marta, 1997 88129 n.r. 1218 Italy Communication Scales (S) No No
b, c
Sabbagh, 1995 212104 b12 04 Canada SSP (O) Yes Yes
Singer et al., 1985 interb 191584 b12 04 USA SSP (O) Yes No
271584 b12 04 USA SSP (O) Yes No
Slobodnik, 1997 8686 b12 N 18 USA PBI Care (S) No No
Snider, 1997 6679 b12 N 18 USA PBI Care (S) No No
Tessier et al., 2005 538 n.r. 12 04 Canada Shortened version AQS (P) No No
Tessier, Tarabulsy, & Moss, 2006b, c 642104 b12 04 Canada SSP (O) Yes Yes
Van Londen, Juffer, & 552104 b12 04 Netherlands SSP (O) Yes Yes
Van IJzendoorn, 2007b, c
Verssimo & Salvaterra, 2006b 1062516 b12 04 Portugal AQS (O) Yes No
Vorria et al., 2006b 612516 12 412 Greece AQS (O) Yes No

Note. Na: number of adoptees; Nc: number of non-adopted controls; n.r.: data were not reported; Age at arrival, m: Age at arrival in months; Age at study, y: Age at study in years;
(O): Observer Report; (P): Parent Report; (S): Self Report; AQS: Attachment Q-sort; AAI: Adult Attachment Interview; SSP: Strange Situation Procedure; PBI: Parental Bonding
Instrument; IPPA: Inventory of Parent and Peer Attachment.
a
The children in this study were compared with the comparisons in an other study.
b
The children were compared with a normative distribution or normative score.
c
Observational studies used in the secondary analyses.

would overrule the inuence of age at assessment. Number of
placements prior to adoption, social economic status of the adoptive
parents as well as prevalence of pre-adoption adversity could not be
used as moderators, because these variables were often not reported
in the pertinent studies. Studies were coded as one of the categories of
a moderator when at least 75% of the sample could be grouped into
that category. Studies in which less than 75% of the sample could be
included in one category of the moderator were coded as mixed.
Furthermore, we examined in the samples with domestically adopted
children whether more than 75% of the children were reported to have
been adopted from foster care. We concluded that there were no
studies that met this criterion (for example, only 5 out of the 106
children in the study of Verssimo and Salvaterra (2006) and none of
the children in the study of Vorria et al. (2003) were adopted from
foster care).
For studies that used the (adapted) SSP and provided information
about the number of children classied as disorganized (e.g., Dozier,
Stovall, Albus, & Bates, 2001; Juffer, Bakermans-Kranenburg, & van
IJzendoorn, 2005), controlling (e.g., Marcovitch et al., 1997) or
disorganized/controlling and insecure-other/Cannot Classify (e.g.,
O'Connor et al., 2003) two effect sizes were derived, one for
attachment security and one for attachment disorganization.
6. Meta-analytic procedures
We conducted ve meta-analyses, four for the core set of studies
with observational measures of attachment and one for the entire set
of adoption studies. For the core set we conducted separate meta-
analyses for adoptive and foster studies, both for attachment security
and attachment disorganization. Some study outcomes could be
directly inserted into Borenstein, Rothstein, and Cohen's (2000)
Comprehensive Meta-Analysis (CMA) program. Other outcomes had
to be re-computed or transformed before they could be inserted into
CMA. For the studies in which the (adapted) SSP was used, we
employed Wilson Effect Size Calculator (Lipsey & Wilson, 2001) to
compare the distributions of the adoption samples with the normative
distribution. Eventually for all studies Cohen's d was computed.
According to Cohen's criteria, ds of 0.20, 0.50, and 0.80 represent
small, moderate, and large effect sizes, respectively (Cohen, 1988).
Because we hypothesized that adopted children would show less
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
Fig. 1. Percentages of disorganized attachment classications based on the Strange
Situation Procedure. Results are reported for adoptees (k = 11), institutionalized children
(k = 2, Vorria et al., 2003; Zeanah et al., 2005), and normative data (k = 15) from Van
IJzendoorn et al. (1999). All percentages differ signicantly from each other at p b .01.
secure and more disorganized attachment (see Introduction) than the
comparison group, the effect sizes of studies presenting outcomes in
this direction were given a positive sign. A negative sign was given to
the outcomes of studies where the adopted children showed more
secure or less disorganized attachment. To examine whether outlying
effects sizes were present, all effect sizes were transformed into
Fisher's Z which were standardized. The study of Golombok, Cook,
Bish, and Murray (1995) had an outlying effect size (d = 2.28) which
exceeded the preset limit of z b 3.29 (Tabachnick & Fidell, 2007).
To avoid an excessive inuence this study was not included in the
meta-analysis.
The effect sizes within one subset of a moderator can be homo-
geneous or heterogeneous. This homogeneity was tested with the Q
statistic (Lipsey & Wilson, 2001). When the effect sizes were
homogeneous, the xed effect parameters were used. Random effect
parameters were used when the effect sizes within one subset were
heterogeneous. To test the inuence of moderators, the Q statistic for
between-group differences was calculated (Borenstein et al., 2000).
This statistic indicates signicant differences between the subgroups
of a moderator. When one or more of the subsets were heterogeneous,
random effect models were used. Fixed models were used when all
subsets were homogeneous. When moderators could not be tested
due to the small size of the subsets, the overlap of the 85% Condence
Intervals (CI) of the subsets was examined (Bakermans-Kranenburg
et al., 2003; Goldstein & Healy, 1995). This provides a global test
of the contrasts between combined effects of subsets grouped by
moderators.
A publication bias may arise when non-signicant ndings
remain unpublished. The results of these possibly missing studies
can be estimated using Duval and Tweedie's trim and ll method
(Duval & Tweedie, 2000a,b). In this method a scatter plot (called a
funnel plot) is created in which the effect sizes of the studies are
plotted against the sample size or standard error. If the plot is shaped
like a funnel no publication bias is present. However, due to the
reduced chance for smaller and non-signicant studies to be
published, studies from the bottom left hand corner may be missing
(the le-drawer problem; Mullen, 1989). With the trim and ll
method these missing studies are lled in and an (adjusted) effect
size is calculated. We also examined the stability of the results. In this
jackknife procedure it is analyzed whether the overall effect size
changes signicantly when the combined effect sizes are calculated
after the successive removal of one effect size (Borenstein et al.,
2000). For each meta-analysis we also calculated the number of
studies with a non-signicant result that would be required to bring
the combined effect size of a meta-analysis to a non-signicant level
(fail-safe number; Mullen, 1989).
415
7. Results
First, a secondary analysis was conducted to compare the distribu-
tions of attachment classications of adopted children with the
normative distribution of non-adopted children. Secondly, the out-
comes of the meta-analyses are presented. We start with the analyses
of the adopted children concerning attachment security and attach-
ment disorganization in the core set of studies (based on observational
assessments), after which the comparison with the foster children is
reported. We conclude with a broad-band meta-analytic approach of
all adoption studies, using all types of measurements.
7.1. Secondary analyses
In order to compare the distributions of attachment classications
of adoptees with attachment distributions of normative, non-adopted
children, the combined attachment distribution of the adopted
sample was calculated. We collected all studies that used the SSP
and reported on the ABCD-classication of adoptees (k = 11 studies).
For an overview of these studies, see Table 1. The distribution of
attachment classication in normative samples is 62%, secure 15%
avoidant, 9% ambivalent, and 15% disorganized (Van IJzendoorn et al.,
1999; N = 2,104). The analyses revealed that the adopted children
showed a signicantly different distribution, 2 (3, N = 2572) = 106.41,
p b .05. Adopted children showed fewer secure attachments, 47%
secure, and more disorganized attachments, 31% disorganized, than
non-adopted children (k = 11, n = 468 adopted children). It should be
noted, however, that the adoptees compared favorably to children in
institutional care regarding disorganized attachment and secure
attachment (see Introduction; 73% disorganized and 11% secure, 2
studies, N = 181; Vorria et al., 2003; Zeanah et al., 2005) (see Fig. 1).
Based on these secondary analyses, which suggested that adoptees
were more often disorganized and showed fewer secure attachments
than their non-adopted counterparts, we examined the attachment
relationships of adoptees in more depth in several meta-analyses.
7.2. Attachment security and disorganization of adoptees in the core set
of studies
7.2.1. Attachment security
The core set consisted of 17 adoption studies measuring attach-
ment security using the SSP or the AQS (reported in 13 publications;
Table 1). Instead of using the classical SSP (Ainsworth et al., 1978), we
also included studies that used the adapted SSP (see Method), such as
the CassidyMarvin system, the MainCassidy system, and Critten-
den's PAA. Before including these studies in the meta-analyses, we
tested if including these measurements inuenced the effect size of
the total core set. The removal of the study using the MainCassidy
system and the exclusion of the two studies using Crittenden's PAA,
Fig. 2. Risk of insecure attachment for adoptees placed before or after their rst birthday
(k = 17).
416 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421

Table 2
Attachment relationships of foster children vs. non-fostered comparisons

Source NfNc Age at arrival, Age at study, Country of Measure Core set Core set
m y study (informant) (in)secure disorganized
Chew, 1998a 72516 12 04 USA AQS (O) Yes No
242516 12 04 USA AQS (O) Yes No
a
Cole, 2005 462104 b12 04 USA SSP (O) Yes Yes
Dozier et al., 2001a 502104 b12 04 USA SSP (O) Yes Yes
Lamb et al., 1985 66 b12 04 USA Adapted SSP (O) Yes No
Oosterman, 2007a; Oosterman, & Schuengel, 2007 612516 12 412 Netherlands AQS (O) Yes No
Ponciano, 2002a 492516 b12 04 USA AQS (O) Yes No
272516 12 04 USA AQS (O) Yes No
Rodning et al., 1991a 112104 b12 04 USA SSP (O) Yes Yes
72104 b12 04 USA SSP (O) Yes Yes
Swanson et al., 2000a 122104 b12 04 USA SSP (O) Yes Yes

Note. Nf: number of foster children; Nc: number of non-fostered controls; n.r.: data were not reported; Age at arrival, m: Age at arrival in months; Age at study, y: Age at study in years;
(O): Observer Report; AQS: Attachment Q-sort; SSP: Strange Situation Procedure.
a
The children were compared with a normative distribution or normative score.

both individually resulted in a (minimal) change of 0.01 for both the
normal and the adjusted effect size. As sufcient studies using the
classical SSP and the CassidyMarvin system were present, we used a
moderator-analysis to test for differences between the effect sizes
of both types of measurements. No signicant differences were found
(Q(1) = 3.17, p N .05). In the total set of adoption studies a modest but
signicant effect size was found (d = 0.34, CI = 0.110.57, n = 722
adoptees) in a heterogeneous set of 17 studies (see Table 3). The
funnel plot showed a publication bias. With the trim-and-ll
procedure, ve studies were trimmed and replaced, resulting in a
non-signicant adjusted effect of d = 0.20 (CI = 0.020.43). This
suggests that adopted children are as securely attached to their
parents as non-adopted controls. The fail-safe number was k = 57. The
jackknife procedure yielded a similar point estimate and the same CIs.
Because of the heterogeneity of the set of studies, several
moderators were tested. As only two adoption studies in the core
set used the AQS, the contrast between AQS and SSP was not tested.
However, the 85% CIs around the point estimate of both subsets did
overlap, suggesting that no differences between the effect sizes of the
subsets were present.
The difference between the effect sizes of the children adopted
before and after 12 months of age was signicant, Q(1) = 15.68, p b .01
(see Fig. 2). Children who were adopted before 12 months of age
showed secure attachments as often as non-adopted children (d = 0.08,
CI = 0.090.25, n = 524 adoptees) in a homogeneous set of 12 studies.
However, children adopted after 12 months of age showed signicantly
less attachment security than non-adopted children in a homogeneous
set of 5 studies, d = 0.80 (CI = 0.491.12, n = 198 adoptees). Continent of
origin was not a signicant moderator, Q(1) = 0.95, p N .05 when all the
European children were grouped together. However, when the Eastern
European adoptees were analyzed separately, continent of origin was a
signicant moderator, Q(1) =5.73, p b .05. Asian adoptees showed the
same level of attachment security as non-adopted children (d = 0.12,
CI = 0.130.36, n = 227 adoptees) in a homogeneous set of 4 studies,
while Eastern European adoptees showed less attachment security
than non-adopted children (d = 0.58, CI = 0.290.87, n = 216 adoptees)
in a homogeneous set of 6 studies. The set of studies with children
adopted from other European countries or from North America was too
small to be included in the analyses. The other moderators, such as
age at assessment, time in family, same- or transracial placement,
international or domestic placement, publication outlet and year of
publication were not signicant.1
7.2.2. Attachment disorganization
We examined whether adopted children were more often
classied as disorganized in the SSP than their non-adopted controls.
Eleven studies reporting on the number of disorganized children were
included (reported in 8 publications, see Table 1).
A signicant positive effect size for disorganized attachment was
found, d = 0.46 (CI = 0.140.77, n = 468 adoptees) in a heterogeneous set
of outcomes. The funnel plot showed some publication bias. With the
trim-and-ll procedure 2 studies were trimmed and replaced,
resulting in a signicant adjusted effects of d = 0.36 (CI = 0.040.68).
This means that more adopted children showed disorganized
attachment compared to their non-adopted counterparts. No sig-
nicant moderators or non-overlapping CIs were found. The fail-safe
number was k = 42. The point estimate and CIs computed with the
jackknife procedure remained the same.2
7.3. Comparison with foster children
Eleven foster studies (reported in 8 publications; Table 2)
measured attachment security using the SSP or the AQS. In this
homogeneous set of 11 studies a non-signicant effect size was found
(d = 0.07, CI = 0.160.30, n = 300 foster children). A publication bias
was indicated in the funnel plot. With the trim-and-ll procedure
three studies were trimmed and replaced, resulting in a non-
signicant adjusted effects of respectively d = 0.06 (CI = 0.270.15).
This means that the foster children are as securely attached to their
foster parents as children reared in their biological family. This
converges with the non-signicant overall effect size found for the
group of adoptees.
Five studies reported on the number of disorganized foster
children (reported in 4 publications, see Table 2). For these studies
an effect size comparable with that of the adopted children was
found, d = 0.41 (CI = 0.070.74, n = 126 foster children) in a homo-
genous set of outcomes. A publication bias was shown in the
funnel plot. With the trim-and-ll procedure 2 studies were
trimmed and replaced, resulting in a signicant adjusted effects of
d = 0.35 (CI = 0.020.67). This means that, again comparable with
adoptees, the foster children showed more disorganized attach-
ment compared to children reared by their biological parents. The
fail-safe number was rather small, k = 5. The combined effect size
computed with the jackknife procedure became non-signicant
when the studies of Dozier et al. (2001) or Cole (2005) were
removed, with adjusted effect sizes of respectively 0.30 (CI = 0.10
0.71) and 0.38 (CI = 0.020.78). The 85% CI intervals for both
attachment security and disorganization in studies with adoptees
versus foster children did overlap, indicating that adopted children
have comparable rates of insecure and disorganized attachment as
foster children.

1 2
Tables with data not presented in the article may be requested from the authors. See footnote 1.
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421 417

Table 3 we found more disorganized attachments among adopted children in

Meta-analytic resultsa of the core set of studies comparing the security of attachment of the core set of studies. When studies using self-report measures, such
adoptees with non-adopted controls
as questionnaires and interviews, were added to the meta-analysis the
k d Na 95% CI Q Q for p effects were no longer signicant. According to this broad-band meta-
contrast analytic approach adoptees had similar attachment relationships with
Total set 17 0.34 772 0.110.57 33.39 their adoptive parents as their non-adopted counterparts. The broad-
band meta-analysis did not reveal any inuence of age at arrival. These
Sample characteristics
Measurement n.a. results suggest that self-report measures may not be as sensitive as
SSP/adapted SSP 15 0.31 555 0.140.48 14.77 observational measures in revealing a clinically meaningful age effect.
AQSb 2 0.19 167 0.400.78 16.29
Age arrival 15.68 .00 8.1. Attachment disorganization
b 12 months 12 0.08 524 0.090.25 16.90
N 12 months 5 0.80 198 0.491.12 0.81
Age assessment 2.68 .10 As expected, based on the presence of risk factors before the
04 years 10 0.18 450 0.080.45 18.71 adoption, adopted children showed more disorganized attachment
412 yearsc 7 0.55 272 0.290.81 6.90 than non-adopted children. This nding may be explained by the
Time in family 1.06 .59
children's experiences of maltreatment and neglect before their
012 months 6 0.13 220 0.120.38 5.89
1324 months 4 0.53 135 0.160.89 1.60 placement in an adoptive family. The ndings for attachment
2574 months 7 0.34 367 0.020.69 22.92 disorganization were independent of age at placement: Both early
Placement 0.48 .49 and later placed adoptees showed more disorganized attachments.
Domestic 4 0.19 243 0.260.64 16.59 This outcome converges with Dozier and Rutter's (2008) suggestion,
International 11 0.32 462 0.130.50 13.77
that children are particularly vulnerable for caregiving experiences
Not reportedd 1 0.90 11 0.622.43
Mixedd 1 0.73 6 1.683.14 during their rst year of life. Experiences of maltreatment, deprivation
Continent of origin 0.95 .62 and neglect during the infant's rst weeks or months may have long-
Europe 9 0.42 432 0.350.82 7.27 lasting consequences for the development of the organization of
Asia 4 0.12 227 0.130.36 4.34
attachment. Thus, not only adoptees placed after their rst birthday,
North Americab 2 0.37 46 0.261.00 0.42
Not reportedd 2 0.86 17 0.432.14 0.01 but also adoptees who have experienced adverse conditions only
Transracial placement 0.59 .44 during their rst months of life may be more prone to develop
Yes 5 0.15 296 0.090.38 5.27 disorganized attachments. Besides institutionalized children, the best
No 10 0.40 459 0.090.70 26.18 comparison group of biological parentchild dyads may be maltreated
Not reportedd 2 0.86 17 0.432.14 0.01
children whose attachment is assessed around their rst birthday.
Note. Na: Number of adoptees. Researchers studied maltreated children as young as 13 months of age
n.a.: not applicable; p b .05; p b .01; p b .001.
a and found high rates of disorganized attachment: 82% to 93% (Carlson,
For trimmed effect sizes see Results section.
b
Subsets with fewer than 4 studies were excluded from the contrast. Cicchetti, Barnett, & Braunwald, 1989; Cicchetti, Rogosch, & Toth,
c
The oldest children in the core set were 7 years of age (placed in the subset 4 2006). The overrepresentation of disorganized attachments in mal-
12 years, see Method). treated children converges with the high rates of disorganized
d
Excluded from the contrast. attachments in institutionalized children (Vorria et al., 2003; Zeanah
et al., 2005). Contrasting the ndings on disorganized attachment of
7.4. Attachment relationships of adoptees: broad-band approach
maltreated and institutionalized children (73% to 93%) with our meta-
analytic ndings for adopted children (31% disorganized attachment)
In the meta-analysis focusing on all types of measurements, 39
suggests that adopted children show an impressive although
studies (reported in 31 publications, see Table 1) measuring the
incomplete catch-up after their placement (31% vs. 15% in normative
attachment relationship of adoptees were included. The comparison
groups).
of the adoptees with the non-adopted controls showed a small but
signicant effect size (d = 0.24, CI = 0.100.37, N = 2912 adoptees) in
8.2. Moderators of attachment security and disorganization
favor of the comparisons, see Table 3. However, the funnel plot
showed a publication bias. With the trim-and-ll procedure eight
Few of the potential moderators appeared to make a signicant
studies were trimmed and replaced, resulting in a non-signicant
difference for attachment security and disorganization. Study char-
adjusted effect of d = 0.12 (CI = 0.020.26). The fail-safe number was
acteristics such as publication outlet, year of publication, and continent
k = 230. The point estimate and CIs computed with the jackknife
of study were not associated with any of the effect sizes. We did nd
procedure remained the same. To account for the heterogeneity of the
that Eastern European children were less often securely attached than
studies in the subset, various moderators were tested. However, no
Asian adoptees. As Eastern European children are suggested to have
signicant moderators were found.3
experienced the most severe deprivation (e.g., Miller, 2005; Rutter et
al., 2004), which may negatively inuence the development of new
8. Discussion
attachment relationships after placement, this outcome was expected.
Unfortunately, we were not able to study the inuence of continent of
Because of the untoward early life experiences that many adopted
origin in combination with the inuence of age at placement and
children endured we expected fewer secure attachments and more
thereby disentangle the inuences of both moderators, as all Asian
attachment disorganization in adopted children compared to non-
children were placed before their rst birthday.
adopted children. Overall, the (trimmed) results of our meta-analyses
Although for example Rosenthal, Groze, Curiel, and Westcott
of a core set of studies with observational attachment measures
(1991) and Singer, Brodzinsky, Ramsay, Steir and Waters (1985)
showed that adopted children were as securely attached as their non-
reported fewer positive parentchild relationships in transracial
adopted counterparts. However, children who were adopted after
adoptees compared to same-race adoptees, we did not replicate
their rst birthday showed signicantly less attachment security than
these results. Attachment ndings were independent of type of
non-adopted children and this effect size was large. As hypothesized,
placement domestic or international and same-race or transracial
placements. However, as Rosenthal et al. (1991) primarily attributed
3
See footnote 1. the differences to differing characteristics at adoptive placement, and
418 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
Singer et al.'s (1985) study was based on a small sample size (n = 19
transracially adopted infants), the results of the meta-analysis are not
unexpected. The fact that parents and children do not share the same
race or the same country of origin may not be essential for the
development of new attachment relationships. The (changed) envir-
onment provided by the parents is probably of more signicance. In
the same vein, similarities between same-race and transracial
adoptees have been found in a meta-analysis on self-esteem (Juffer
& Van IJzendoorn, 2007).
Age at assessment and time in the family were not signicant
moderators either. We had hypothesized that children would need
some time before they could prot from the new family environment.
However, as all but one of the studies examining children who had
lived in their adoptive family for less than a year reported on children
who had lived with their new family for at least eight months, these
results are not that surprising. Eight months may be sufcient time to
develop a secure attachment relationship with the new parents.
Moreover, with the exception of one study, all studies examining
children who had lived in their adoptive family for less than a year
reported on early placed children, who are expected to develop secure
attachment relationships as often as non-adopted children. In the
same vein, Stovall-McClough and Dozier (2004) reported that early
placed foster children already began to show secure attachment
behaviors within the rst two months of their placement.
As expected, age at placement was a signicant moderator for
attachment security in adoptees, with early placed adoptees showing
secure attachments as often as non-adopted children. Children who
are placed before their rst birthday may have experienced depriva-
tion for shorter periods of time than later placed adoptees, resulting in
a more normative development of attachment relationships (Bowlby,
1982). Alternatively, it may be easier for early placed children to
become securely attached because they are placed with new parents
and receive sensitive care in a stage when attachment is still
developing (Ainsworth et al., 1978; Bowlby, 1982). It may be easier
to prevent insecure attachment than to change insecure attachment.
8.3. Comparison with foster children
To compare adoptees with foster children we conducted additional
meta-analyses for foster children's attachment security and disorga-
nization. The effect sizes for attachment security of the adopted and
foster children were comparable, as were the effect sizes for
attachment disorganization (adoptees, d = 0.36; foster children,
d = 0.35). As mentioned above, the high rate of disorganized attach-
ment of the adopted children may be explained by the inuence of the
adverse circumstances these children experienced before their
placement. The same might be true for the foster children, as many
of them have also experienced maltreatment and/or neglect before
placement (e.g., Chernoff et al., 1994).
8.4. Limitations
Although publication bias was present in all sets of studies, most
effect sizes remained signicant after correction for such a bias
through the trim-and-ll procedure. The effect size for the total set of
studies on attachment relationships was small before trimming
(d = 0.24), and lost its signicance after trimming. Similarly, in the
core set, the effect size for attachment security in the adoptive group
was not signicant after trimming. The fact that studies had to be
trimmed and lled may point to a le-drawer problem (Mullen, 1989),
suggesting that non-signicant results in this eld are not published
as much as signicant outcomes. Rosenthal (1991, p. 106) suggested
that a fail-safe number of 5k + 10 (k = number of studies included) is a
general criterion for robustness. This criterion was not achieved for
any of the sets, suggesting that the outcomes of our meta-analyses
must be interpreted with some caution.
Foster care arrangements may differ in the USA and Europe (for
example offering adoption from foster care or not; Jones, 2008;
Warman & Roberts, 2003). One of the limitations of our meta-analyses
is that only one of the foster care studies was conducted outside of the
USA (Oosterman, 2007). With the jackknife procedure (see Method)
the meta-analytic outcomes were similar when this specic foster
care study was removed from the analyses.
Because a relatively limited number of studies were available for
our meta-analysis, we could only examine broad categories for
example, continents of origin instead of separate countries and
within the broad categories contrasting study outcomes may remain
hidden until more primary studies become available. A risk factor like
deprivation or pre-placement adversity is an important predictor of
child development. Unfortunately, this moderator could not be
included in the meta-analyses since in many studies insufcient
information about the care background of the children was reported.
Moreover, important details of the children's caregiving history could
not be taken into account because they were unknown in many
studies (e.g., number of placements). Similarly, we could not include
adoptive parents' sensitivity or parenting behavior, or their attach-
ment representation.
In our meta-analyses we have compared children adopted before
and after one year of age. It would be interesting to distinguish more
subgroups, for example children adopted between one and two years
of age, and children adopted after two years of age. Unfortunately, the
set of observational studies including children adopted after their rst
birthday was too small to conduct this analysis.
8.5. Clinical implications
Interventions in adoptive families may be needed to support
parents' sensitivity and enhance adopted children's attachment
security (Juffer et al., 2008). A meta-analysis of intervention studies
showed that interventions that successfully increase parental sensi-
tivity are also successful in enhancing attachment security. Further-
more, a doseresponse relation was revealed: interventions with
larger effects on sensitivity resulted in larger effects on attachment
security (Bakermans-Kranenburg et al., 2003). An intervention aimed
at promoting adoptive parents' sensitivity not only resulted in
increased maternal sensitivity but also in a reduced number of
disorganized attachments (Juffer et al., 2005). However, the children
in this study were placed at a very early age (M = 10 weeks) and we do
not know whether these ndings can be generalized to (somewhat)
older placed children. Comparably, Stovall and Dozier (2000), using
detailed diaries to study the development of attachment in foster
children, concluded that foster parents of late placed foster children
not only need to be sensitive to promote secure attachments, but also
need to provide therapeutic caregiving by challenging the foster
children's alienating behavior. Promising intervention studies in foster
families targeting children's attachment behavior as well as their
biobehavioral stress regulation point to positive effects on attachment
security and neurobiological adaptation (Dozier, 2003; Dozier, Albus,
Fisher, & Sepulveda, 2002; Dozier, Higley, Albus, & Nutter, 2002;
Fisher, Gunnar, Dozier, Bruce, & Pears, 2006). Further research on the
inuence of parenting behavior on children's attachment behavior
after the adoptive placement may reveal new insights into how
adopted children become securely attached. In addition, intervention
studies with adoptive families may show how insecure attachment
strategies can be changed, how long this process usually takes and
which behavioral and neurobiological mechanisms can be held
responsible for recovery.
9. Conclusion
In conclusion, this meta-analysis suggests that adopted children
can overcome early adversity and risks and form secure attachments
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
as often as their normative counterparts. The same was true of foster
children. These outcomes lend support to Bowlby's (1952, 1988)
hypothesis that corrective attachment experiences may enhance
attachment security. But this catch-up is not without limits: Children
who are adopted after their rst birthday are less capable of
developing secure attachments. Moreover, the adoptees show
disorganized attachments more often than their normative peers,
and again we found comparable outcomes in foster children. Adopted
children are however considerably less often disorganized than
institutionalized children (Vorria et al., 2003; Zeanah et al., 2005).
Therefore, adoption may be seen as an effective intervention (Juffer &
Van IJzendoorn, 2006), offering children who lack the care of
their birth parents the chance to develop more secure attachment
relationships.
References4
Adoption and Foster Care Analysis and Reporting System (2008). The AFCARS Report.
Preliminary FY 2006 Estimates as of January 2008 (14). Retrieved September 2,
2008 from http://www.acf.hhs.gov/programs/cb/stats_research/afcars/tar/
report14.pdf
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A
psychological study of the strange situation. Hillsdale, N.J.: Erlbaum.
Armsden, G. C., & Greenberg, M. (1987). The inventory of parent and peer attachment:
Individual differences and their relationship to psychological well-being in
adolescence. Journal of Youth and Adolescence, 16, 427454.
Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Less is more:
Meta-analyses of sensitivity and attachment interventions in early childhood.
Psychological Bulletin, 129, 195215.
Bartel, T.M.C., (2006). Factors associated with attachment in international adoption.
Dissertation Abstracts International, 66 (12), 4291A (UMI No. 3203835).
Borders, L. D., Penny, J. M., & Portnoy, F. (2000). Adult adoptees and their friends:
Current functioning and psychosocial well-being. Family Relations, 49, 407418.
Borenstein, M., Rothstein, D., & Cohen, J. (2000). Comprehensive meta-analysis: A
computer program for research synthesis [Computer software]. Englewood, NJ:
Biostat.
Bowlby, J. (1952). Maternal care and mental health, 2nd ed. Geneva: WHO.
Bowlby, J. (1982 2nd ed.). Attachment and loss (Vol. 2). Attachment. New York: Basic
Books.
Bowlby, J. (1973). Attachment and loss (Vol. 2). Separation: Anxiety and anger. New York:
Penguin Books.
Bowlby, J. (1988). A secure base. Clinical applications of attachment theory London:
Routledge.
Brooks, B., Simmel, C., Wind, L., & Barth, R. P. (2005). Contemporary adoption in the
United States: Implications for the next wave of adoption, theory, research, and
practice. In D. M. Brodzinsky & J. Palacios (Eds.), Psychological Issues in Adoption:
Research and Practice (pp. 146). Westport, CT: Praeger Publishers.
Bretherton, I., & Munholland, K. A. (1999). Internal working models in attachment: A
construct revisited. In J. Cassidy & P. Shaver (Eds.), Handbook of attachment: Theory,
research, and clinical applications (pp. 89111). New York: Guilford.
Brown, L.E., (2000). The role of perceived similarity to parents in adopted children's
adjustment. Dissertation Abstracts International, 62 (01), 538B (UMI No. 3000886).
Carlson, V., Cicchetti, D., Barnett, D., & Braunwald, K. (1989). Disorganized/ disoriented
attachment relationships in maltreated infants. Developmental Psychology, 25,
525531.
Caspers, K., Yucuis, M. S. W., Troutman, B., Arndt, S., & Langbehn, D. (2007). A sibling
adoption study of adult attachment: The inuence of shared environment on
attachment states of mind. Attachment & Human Development, 9, 375391.
Cassidy, J., & Marvin, R. S. (1992). Attachment organization in preschool children:
Procedures and coding manual. Unpublished manuscript, MacArthur Working Group
on Attachment, Seattle, WA.
Cassidy, J., & Shaver, P. (Eds.). (1999). The handbook of attachment: Theory, research and
clinical applications. New York: Guilford Press.
Castle, J., Groothues, C., Bredenkamp, D., Becket, C., O'Connor, T., Rutter, M., et al. (1999).
Effects of qualities of early institutional care on cognitive attainment. American
Journal of Orthopsychiatry, 40, 424437.
Chew, R.J., (1998). Development, attachment and play in drug-exposed children of
multiple foster care placements. Dissertation Abstracts International, 57 (06),
3090B (UMI No. 9835994).
Chisholm, K. (1998). A three year follow-up of attachment and indiscriminate
friendliness in children adopted from Romanian orphanages. Child Development,
69, 10921106.
Chisholm, K., Carter, M. C., Ames, E. W., & Morison, S. J. (1995). Attachment security and
indiscriminately friendly behavior in children adopted form Romanian orphanages.
Development and Psychopathology, 7, 283294.
Chernoff, R., Combs-Orme, T., Risley-Curtiss, C., & Heisler, A. (1994). Assessing the health
status of children entering foster care. Pediatrics, 93, 594601.
4
References marked with an asterisk indicate studies included in the meta-analyses.
419
Chugani, H. T., Behen, M. E., Muzik, O., Juhasz, C., Nagy, F., & Chugani, D. C. (2001). Local
brain functional activity following early deprivation: A study of postinstitutiona-
lized Romanian orphans. Neuroimage, 14, 12901301.
Cicchetti, D., & Barnett, D. (1991). Attachment organization in maltreated preschoolers.
Development and Psychopathology, 3, 397411.
Cicchetti, D., Rogosch, F. A., & Toth, S. (2006). Fostering secure attachment in infants in
maltreating families through preventive interventions. Development and Psycho-
pathology, 18, 326649.
Cohen, J. (1988). Statistical power analysis for the behavioural sciences (rev. edn).
New York: Academic Press.
Cole, S. A. (2005). Infants in foster care: Relational and environmental factors affecting
attachment. Journal of Reproductive and Infant Psychology, 23, 4361.
Crittenden, P. M. (1992). Preschool assessment of attachment. Miami, FL: Unpublished
manuscript, Family Relations Institute.
Crittenden, P. M., & Ainsworth, M. D. S. (1989). Child maltreatment and attachment
theory. In D. Cicchetti & V. Carlson (Eds.), Child maltreatment: Theory and research on
the causes and consequences of child abuse and neglect (pp. 432463). New York:
Cambridge University Press.
Crittenden, P. M., Claussen, A. H., & Kozlowska, K. (2007). Choosing a valid assessment of
attachment for clinical use: A comparative study. Australian & New Zealand Journal
of Family Therapy, 28, 7887.
Department for Children, Schools and Families (2008). Statistical Volume 20052006.
Retrieved April 18, 2008 from http://www.dfes.gov.uk/rsgateway/DB/VOL/
v000721/All_National_Tables.xls
Dozier, M. (2003). Attachment-based treatment for vulnerable children. Attachment and
Human Development, 5, 253257.
Dozier, M., Albus, K., Fisher, P. A., & Sepulveda, S. (2002). Interventions for foster
parents: Implications for developmental theory. Development and Psychopathology,
14, 843860.
Dozier, M., Higley, E., Albus, K., & Nutter, A. (2002). Intervening with foster care infants'
caregivers: Targeting three critical needs. Infant Mental Health Journal, 23, 541554.
Dozier, M., & Rutter, M. (2008). Challenges to the development of attachment
relationships faced by young children in foster and adoptive care. In J. Cassidy &
P. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications
(pp. 698717)., (2nd ed.) New York: Guilford.
Dozier, M., Stovall, K. C., Albus, K. E., & Bates, B. (2001). Attachment for infants in foster
care: The role of caregiver state of mind. Child Development, 72, 14671477.
Duval, S., & Tweedie, R. (2000). A nonparametric trim and ll method of accounting for
publication bias in meta-analysis. Journal of the American Statistical Association, 95, 8998.
Duval, S., & Tweedie, R. (2000). Trim and ll: A simple funnel-plot-based method of
testing and adjusting for publication bias in meta-analysis. Biometrics, 56, 455463.
Egeland, B., & Sroufe, L. A. (1981). Attachment and early maltreatment. Child
Development, 52, 4452.
Farina, L., Leifer, M., & Chasnoff, I. J. (2004). Attachment and behavioural difculties in
internationally adopted Russian children. Adoption & Fostering, 28(2), 3849.
Feeney, J. A., Passmore, N. L., & Peterson, C. C. (2007). Adoption, attachment, and
relationship concerns: A study of adult adoptees. Personal Relationships, 14, 129147.
Finzi, R., Ram, A., Har-Even, D., Shnit, D., & Weizman, A. (2001). Attachment styles and
aggression in physically abused and neglected children. Journal of Youth and
Adolescence, 30, 769876.
Fischman, D.A., (1996). Assessment of object relations and attachment capacity in adult
adoptees. Dissertation Abstracts International, 57 (02), 1500B (UMI No. 9618658).
Fisher, P. A., Gunnar, M. R., Dozier, M., Bruce, J., & Pears, K. C. (2006). Effects of
therapeutic interventions for foster children on behavioral problems, caregiver
attachment, and stress regulatory neural systems. Annals of the New York Academy
of Sciences, 1094, 215225.
Fletcher, J.F., (1995). Correlates of psychosocial adjustment among adult adopted
children. Dissertation Abstracts International, 58 (01), 0290A (UMI No. 9717722).
Fries, A. B. W., Ziegler, T. E., Kurian, J. R., Jacoris, S., & Pollak, S. D. (2005). Early experience
in humans is associated with changes in neuropeptides critical for regulating social
behavior. Proceedings of the National Academy of Sciences, 102, 1723717240.
George, C. (1996). A representational perspective of child abuse and prevention: Internal
working models of attachment and caregiving. Child Abuse & Neglect, 20, 411424.
George, C., Kaplan, N., & Main, M. (1985). Adult attachment interview. : Unpublished
Manuscript University of Carolina, Berkeley.
Golombok, S., Cook, R., Bish, A., & Murray, C. (1995). Families created by the new
reproductive technologies: Quality of parenting and social and emotional
development of the children. Child Development, 66, 285298.
Goldstein, H., & Healy, M. J. R. (1995). The graphical presentation of a collection of
means. Journal of the Royal Statistical Society, Series A (Statistics in Society), 158,
175177.
Gunnar, M. R., Bruce, J., & Grotevant, H. D. (2000). International adoption of
institutionally reared children: Research and policy. Development and Psychopathol-
ogy, 12, 677693.
Gunnar, M. R., & Kertes, D. A. (2005). Prenatal and postnatal risks to neurobiological
development in internationally adopted children. In D. M. Brodzinsky & J. Palacios
(Eds.), Psychological issues in adoption (pp. 4765). London: Praeger.
Gunnar, M. R., Morison, S. J., Chisholm, K., & Schuder, M. (2001). Salivary cortisol levels
in children adopted from Romanian orphanages. Development and Psychopathology,
13, 611628.
Hayward, R. A., & DePanlis, D. (2007). Foster children with an incarcerated parent:
Predictors of reunication. Children and Youth Services Review, 29, 13201334.
Hodges, J., Steele, M., Hillman, S., Henderson, K., & Kaniuk, J. (2005). Change and continuity
in mental representations of attachment after adoption. In D. Brodzinsky & J. Palacios
(Eds.), Psychological Issues in Adoption: Research and Practice (pp. 93116). Westport,
CT: Praeger Publishers.
420 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
Howe, D. (2005). Child abuse and neglect: Attachment, development and intervention.
Basingstoke: Palgrave Macmillan.
Irhammar, M., & Bengtsson, H. (2004). Attachment in a group of adult international
adoptees. Adoption Quarterly, 8(2), 125.
Jensen, H.S., (2004) Adult adoptee attachment and psychopathology: Impact of age at
adoption and disclosure, preplacement history, and family communication.
Dissertation Abstracts International, 65 (04), 2097B (UMI No. 3129157).
Judge, S. (2004). Adoptive families: The effects of early relational deprivation in children
adopted from Eastern European orphanages. Journal of Family Nursing, 10, 338356.
Johnson, K. A. (2004). Wanting a daughter, needing a son: abandonment, adoption, and
orphanage care in China. Yeong & Yeong Book Company: St. Paul, Minnesota.
Jones, J. (2008). Adoption experiences of women and men and demand for children to adopt
by women 1844 years of age in the United States, 2002 (DHHS Publication No. PHS
20081979). Washington, DC: U.S. Government Printing Ofce.
Joseph, S.R., (2002). The relationship between adopted young adults and their non-
adopted silings in the perceptions of current family relationships. Dissertation
Abstracts International, 63 (03), 1550B (UMI No. 3045686).
Juffer, F., Bakermans-Kranenburg, M. J., & Van IJzendoorn, M. H. (2005). The
importance of parenting in the development of disorganized attachment: Evidence
from a preventive intervention study in adoptive families. Journal of Child
Psychology and Psychiatry, 46, 263274.
Juffer, F., Bakermans-Kranenburg, M.J., & Van IJzendoorn, M.H., (2008). Promoting
positive parenting: An attachment-based intervention. London: Lawrence Erlbaum/
Taylor & Francis.
Juffer, F., & Van IJzendoorn, M. H. (2006). The Emanual Miller memorial lecture 2006:
Adoption as intervention. Meta-analytic evidence for massive catch-up and
plasticity in physical, socio-emotional, and cognitive development. Journal of
Child Psychology and Psychiatry, 47, 12281245.
Juffer, F., & Rosenboom, L. G. (1997). Infantmother attachment of internationally
adopted children in the Netherlands. International Journal of Behavioral Develop-
ment, 20, 93107.
Juffer, F., & Van IJzendoorn, M. H. (2005). Behavior problems and mental health referrals
of international adoptees: A meta-analysis. JAMA The Journal of the American
Medical Association, 293, 25012515.
Juffer, F., & Van IJzendoorn, M. H. (2007). Adoptees do not lack self-esteem: A meta-
analysis of studies on self-esteem of transracial, international, and domestic
adoptees. Psychological Bulletin, 133, 10671083.
Kaniuk, J., Steele, M., & Hodges, J. (2004). Report on a longitudinal research project, exploring
the development of attachments between older, hard-to-place children and their
adopters over the rst two years of placement. Adoption and Fostering, 28(2), 6167.
Lamb, M. E., Gaensbauer, T. J., Malkin, C. M., & Schultz, L. A. (1985). The effects of child
maltreatment on security of infantadult attachment. Infant Behavior and
Development, 8, 3445.
Leathers, S. J. (2003). Parental visiting, conicting allegiances, and emotional and
behavioral problems among foster children. Family Relations, 52, 5363.
Lewit, M. E. (1993). Children in foster care. The Future of Children, 3, 192200.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Newbury Park, CA: Sage.
Lis, S. (2000). Characteristics of attachment behavior in institution-reared children. In P.
McKinsey Crittenden & A. Hartl Claussen (Eds.), The organization of attachment rela-
tionships: Maturation, culture and context (pp. 141170). Cambridge: University Press.
Lyons-Ruth, K., Easterbrooks, A., & Cibelli, C. (1997). Infant attachment strategies, infant
mental lag, and maternal depressive symptoms: Predictors of internalizing and
externalizing problems at age 7. Developmental Psychopathology, 33, 681692.
Main, M., & Cassidy, J. (1988). Categories of response to reunion with the parent at age 6:
Predictable form infant attachment classications and stable over a 1-month
period. Development Psychology, 24, 112.
Main, M., & Hesse, E. (1990). Parents' unresolved traumatic experiences are related to
infant disorganized attachment status: Is frightened and/or frightening parental
behavior the linking mechanism? In M. T. Greenberg, D. Cicchetti, & E. M. Cummings
(Eds.), Developmental Psychology, Attachment in the preschool years: Theory, research,
and intervention (pp. 161-18). Chicago and London: University of Chicago Press.
Marcovitch, S., Goldberg, A., Gold, A., Washington, J., Wasson, C., Krekewich, K., et al.
(1997). Determinants of behavioural problems in Romanian children adopted in
Ontario. International Journal of Behavioral Development, 20, 1731.
McGinn, M.F., (2001). Attachment and psychological separation in adopted and
non-adopted infants. Dissertation Abstracts International, 62 (02), 4488B (UMI
No. 3006167).
Millham, J.D., (2003). Institutionalization and attachment of adopted children and pa-
renting stress. Dissertation Abstracts International, 64 (01), 426B (UMI No. 3077443).
Miller, L. C. (2005). The handbook of international adoption medicine. A guide for
physicians, parents, and providers. Oxford: Oxford University Press.
Morison, S. J., Ames, E. W., & Chisholm, K. (1995). The development of children adopted
from Romanian orphanages. Merrill-Palmer Quarterly, 41, 411430.
Moss, E., Bureau, J., Cyr, C., Mongeau, C., & St-Laurent, D. (2004). Correlates of
attachment at age 3: Construct validity of the preschool attachment classication
system. Developmental Psychology, 40, 323334.
Mullen, B. (1989). Advanced basic meta-analysis. Hillsdale, NJ: Erlbaum.
Mller, M., Gibbs, P., & Ariely, S. G. (2002). Predictors of psychological functioning and
adoption experiences in adults searching for their birthparents. Adoption Quarterly,
5(3), 2553.
Nickman, S., Roseneld, A., Fine, P., Macintyre, J. C., Pilowsky, D. J., Howe, R. A., et al.
(2005). Children in adoptive families: Overview and update. Journal of the American
Academy of Child and Adolescent Psychiatry, 44, 987995.
O'Connor, T. G., Marvin, R. S., Rutter, M., Olrick, J. T., & Britner, P. A. (2003). Childparent
attachment following early institutional deprivation. Development and Psycho-
pathology, 15, 1938.
Ongari, B., & Tomasi, F. (2006, July). Attachment patterns and risk factors in adoption. Poster
presented at the meeting of the World Association for Infant Mental Health. Paris: France.
Oosterman, M. (2007). Attachment to foster parents. Amsterdam: Vrije Universiteit
Amsterdam.
Oosterman, M., & Schuengel, C. (2007). Autonomic reactivity of children to separation
and reunion with foster parents. Journal of the American Academy of Child and
Adolescent Psychiatry, 46, 11961203.
Pace, C. S., Messina, S., Zavattini, G. C., & Santona, A. (July, 2006). Diagnosis about
attachment relationships between mothers and children in adoptive families
during the rst eight months of adoption. Poster session presented at the meeting of
the World Association for Infant Mental Health. Paris: France.
Palacios, J., & Snchez-Sandoval, Y. (2005). Beyond adopted/nonadopted comparisons.
In D. M. Brodzinsky & J. Palacios (Eds.), Psychological issues in adoption. Research and
practice (pp. 117144). Westport, CT: Praeger.
Paperny, R.D., (2004). Adoption and attachment: A study of attachment in young adult
adoptees. Dissertation Abstracts International, 65 (01), 466B (UMI No. NQ87067).
Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument. British
Journal of MedicalPsychology, 52, 110.
Passmore, N. L., Fogarty, G. J., Bourke, C. J., & Bakers-Evans, S. F. (2005). Parental
bonding and identity style as correlates of self-esteem among adult adoptees.
Family Relations, 54, 523534.
Placek, P. (2007). National adoption data. In T. C. Atwood, L. A. Allen, V. C. Ravenel, & N. F.
Callahan (Eds.), Adoption Factbook IV. The most Comprehensive Source for Adoption
Statistics Nationwide (pp. 369). Washington: National Council for Adoption.
Ponciano, L.H., (2002). Attachment relationships in foster care: Contributions of the
child, the foster mother and the foster home. Dissertation Abstracts International,
62 (11), 5411B (UMI No. 3032842).
Quinton, D., Rushton, A., Dance, C., & Mayes, D. (1998). Joining new families: A study of
adoption and fostering in middle childhood. Chichester: Wiley & Sons.
Rodning, C., Beckwith, L., & Howard, J. (1991). Quality of attachment and home
environments in children prenatally exposed to PCP and cocaine. Development and
Psychopathology, 31, 351366.
Rosenthal, D., Wender, P. H., Kety, S. S., Schulsinger, F., Welner, J., & Rieder, R. O. (1975).
Parentchild relationships and psychopathological disorder in the child. Archives
General Psychiatry, 32, 466476.
Rosenthal, J. A., Groze, V., Curiel, H., & Wescott, P. A. (1991). Transracial and inracial
adoption of special needs children. Journal of Multicultural Social Work, 1(3), 1331.
Rosnati, R., & Marta, E. (1997). Parentchild relationship as a protective factor in
preventing adolescents' psychosocial risk in inter-racial adoptive and non-adoptive
families. Journal of Adolescence, 20, 617631.
Rosenthal, R. (1991). Meta-analytic procedures for social research. New York: Sage.
Rutgers, A. H., Van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., & Swinkels, S. H. N.
(2007). Autism and attachment: The attachment Q-sort. Autism: The International
Journal of Research and Practice, 11, 187200.
Rutter, M. (1990). Psychosocial resilience and protective mechanisms. In J. Rolf, A. S.
Masten, D. Cichetti, K. H. Nuechterlein, & S. Weintraub (Eds.), Risk and protective
factors in the development of psychopathology (pp. 181214). Cambridge, England:
Cambridge University Press.
Rutter, M. (2005). Adverse preadoption experiences and psychological outcomes. In
D. M. Brodzinsky & J. Palacios (Eds.), Psychological issues in adoption (pp. 6792).
London: Praeger.
Rutter, M. (2006). Genes and behavior. Nature-nurture interplay explained. Oxford:
Blackwell.
Rutter, M., & O'Connor, T. G.& the English and Romanian Adoptees (ERA) Study Team.
(2004). Are there biological programming effects for psychological development?
Findings from a study of Romanian adoptees. Developmental Psychology, 40, 8194.
Sanchirico, A., & Jablonka, K. (2000). Keeping foster children connected to their
biological parents: The impact of foster parent training and support. Child and
Adolescent Social Work Journal, 17, 185203.
Selman, P. (2006). Trends in intercountry adoption: Analysis of data from 20 receiving
countries. Journal of Population Research, 23, 183204.
Singer, L. M., Brodzinsky, D. M., Ramsay, D., Steir, M., & Waters, E. (1985). Motherinfant
attachment in adoptive families. Child Development, 56, 15431551.
Sabbagh, R., (1995). Attachment and behavior toward strangers in Romanian
preschoolers adopted into Canadian families. Unpublished master's thesis,
University of Toronto, Toronto, Ontario, Canada.
Slobodnik, N.J., (1997) Communication about adoption, personality development and
attachment in adoptive families. Dissertation Abstracts International, 58 (03), 1113A
(UMI No. 9724535).
Smyke, A. T., Koga, S. F., Johnson, D. E., Fox, N. A., Marshall, P. J., Nelson, C. A., et al. (2007).
The caregiving context in institution reared and family reared infants and toddlers
in Romania. Journal of Child Psychology and Psychiatry, 48, 210218.
Snider, E.A., (1997) The female adoptee's identity and the post-reunion relationship
with her birth mother. Dissertation Abstracts International, 58 (03), 1548B (UMI
No. 9727660).
Solomon, J., & George, C. (1999). The measure of attachment security in infancy and
childhood. In J. Cassidy & P. Shaver (Eds.), Handbook of attachment: Theory, research,
and clinical applications (pp. 287318). New York: Guilford Press.
Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, W. A. (2005). The development of the
person: The Minnesota study of risk and adaptation from birth to adulthood. New York:
Guilford Press.
Stovall, K. C., & Dozier, M. (2000). The development of attachment in new relationships: Single
subject analyses for 10 foster infants. Development and Psychopathology, 12, 133156.
Stovall-McClough, K. C., & Dozier, D. (2004). Forming attachments in foster care: Infant
attachment behaviors during the rst 2 months of placement. Development and
Psychopathology, 16, 253271.
 L. van den Dries et al. / Children and Youth Services Review 31 (2009) 410421
Swanson, K., Beckwith, L., & Howard, J. (2000). Intrusive caregiving and quality of
attachment in prenatally drugs-exposed toddlers and their primary caregivers.
Attachment & Human Development, 2, 130148.
Tabachnick, B. G., & Fidell, L. S., (2007). Using multivariate statistics, 5th ed. Boston: Allyn
& Bacon.
Takayama, J. I., Wolfe, E., & Coulter, K. P. (1998). Relationship between reason for
placement and medical nding among children in foster care. Pediatrics, 101,
201207.
Tessier, R., Tarabulsy, G. M., & Moss, E. (July, 2006). Age at adoption and adoptive
parental care as key determinants of socio emotional development of internation-
ally adopted children. Paper presented at the meeting of the World Association for
Infant Mental Health. Paris: France.
Tessier, R., Larose, S., Moss, E., Nadeau, L., & Tarabulsy, G. M. (2005). International
adoption in Quebec form 1985 to 2002. Social adjustment in children of international
origin adopted by Quebec families. Quebec, Canada: Laval University.
Testa, M. F. (2004). When children cannot return home: Adoption and guardianship. The
Future of Children, 14, 115129.
Teti, D. M., & Gelfand, D. M. (1997). The preschool assessment of attachment: Construct
validity in a sample of depressed and nondepressed families. Development and
Psychopathology, 9, 517536.
Tirella, L. G., Chan, W., Cermak, S. A., Litvinova, A., Salas, K. C., & Miller, L. C. (2008). Time
use in Russian baby homes. Child: Care, Health and Development, 34, 7786.
Tizard, B., & Rees, J. (1975). The effect of early institutional rearing on the behaviour
problems and affectional relationships of four-year-old children. Journal of Child
Psychology and Psychiatry, 16, 6173.
Van IJzendoorn, M. H., Goldberg, S., Kroonenberg, P. M., & Frenkl, O. J. (1992). The
relative effects of maternal and child problems on the quality of attachment: A
meta-analysis of attachment in clinical samples. Child Development, 63, 840858.
Van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., & Juffer, F. (2007). Plasticity of
growth in height, weight, and head circumference: Meta-analytic evidence of
massive catch-up after international adoption. Journal of Developmental and
Behavioral Pediatrics, 28, 334343.
Van IJzendoorn, M. H., Juffer, F., & Klein Poelhuis, C. W. (2005). Adoption and cognitive
development: A meta-analytic comparison of adopted and non-adopted children's
IQ and school performance. Psychological Bulletin, 131, 301316.
Van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999).
Disorganized attachment in early childhood: Meta-analysis of precursors, con-
comitants, and sequelae. Development and Psychopathology, 11, 225249.
Van IJzendoorn, M. H., Vereijken, C. M. J. L., Bakermans-Kranenburg, M. J., & Riksen-
Walraven, J. M. A. (2004). Assessing attachment security with the Attachment Q-
Sort: Meta-analytic evidence for the validity of the observer AQS. Child Develop-
ment, 75, 11881213.
421
Van Londen, W. M., Juffer, F., & Van IJzendoorn, M. H. (2007). Attachment, cognitive and
motor development in adopted children: Short-term outcomes after international
adoption. Journal of Pediatric Psychology, 32, 12591263.
Verssimo, M., & Salvaterra, F. (2006). Maternal scripts and children's attachment
security in an adopted sample. Attachment & Human Development, 8, 261273.
Vorria, P., Papaligoura, Z., Dunn, J., van IJzendoorn, M. H., Steele, H., Kontopoulou, A.,
et al. (2003). Early experiences and attachment relationships of Greek infants
raised in residential group care. Journal of Child Psychology and Psychiatry, 44,
12081220.
Vorria, P., Papaligoura, Z., Saradou, J., Kopakaki, M., Dunn, J., van IJzendoorn, M. H., et
al. (2006). The development of adopted children after institutional care: A follow-
up study. Journal of Child Psychology and Psychiatry, 47, 12461253.
Warman, A., & Roberts, C. (2003). Adoption and looked after children: An international
comparison. Oxford, UK: Oxford Centre for Family Law and Policy, University of
Oxford.
Waters, E., & Deane, K. (1985). Dening and assessing individual differences in
attachment relationships: Q-methodology and the organization of behavior in
infancy and early childhood. In I. Bretherton, & E. Waters (Eds.), Growing points of
attachment theory and research. Monographs of the Society for Research in Child
Development, vol. 50 (pp. 4165). Serial No. 209.
Weineld, N. S., Sroufe, L. A., Egeland, B., & Carlson, E. (1999). The nature of individual
differences in infantcaregiver attachment. In J. Cassidy & P. Shaver (Eds.), Hand-
book of Attachment: Theory, research, and clinical application (pp. 6888). New York:
Guilford Press.
Werner, E. (2000). Protective factors and individual resilience. In J. P. Shonkoff & S. J.
Meisels (Eds.), Handbook of early childhood intervention (pp. 115132). 2nd ed.
Cambridge, England: Cambridge University Press.
Youngblade, L., & Belsky, J. (1990). Social and emotional consequences of child
maltreatment. In R. T. Ammerman & M. Hersen (Eds.), Children at Risk (pp. 109146).
New York: Plenum.
Zeanah, C. H. (2000). Disturbances of attachment in young children adopted from
institutions. Development and Behavioral Pediatrics, 21, 230236.
Zeanah, C. H., Smyke, A. T., Koga, S., & Carlson, E.& the BEIP Core Group. (2005).
Attachment in institutionalized and community children in Romania. Child
Development, 76, 10151028.
Zeanah, C. H., Scheeringa, M., Boris, N. W., Heller, S. S., Smyke, A. T., & Trapani, J. (2004).
Reactive attachment disorder in maltreated toddlers. Child Abuse and Neglect: The
International Journal, 28, 877888.